Assessment of occupational exposures is a crucial factor in evaluating dose-response relationships and most studies conducted by the Branch have an extensive exposure assessment component. Major assessment efforts in cohort studies have involved exposures to pesticides, benzene, polycyclic aromatic hydrocarbons, and diesel exhaust fumes, and to a broad range of exposures in a cohort study of women in Shanghai and shipyard workers. Cross-sectional studies are being conducted to study the effect of benzene, trichloroethylene, and atrazine on genotoxic, immune, and hormonal parameters. In the case-control design, jobs have been evaluated for a wide variety of exposures, including solvents (including chlorinated hydrocarbons, aromatic, and other solvents), metals (arsenic, chromium, cadmium, lead), electromagnetic fields, polychlorinated biphenyls, dusts (wood dust and other types) , asbestos, formaldehyde, physical activity, nitrosoamines, polycyclic aromatic hydrocarbons, gasoline and diesel exhausts, aromatic amines, and pesticides. These exposures have been evaluated in studies of cancer of the bladder, brain, breast, kidney, lung, larynx, and the lymphatic and hemtopoietic system, and of childhood brain cancer, germ cell tumor cancer, and leukemia Ewings sarcoma and possible associations with parental occupations. Methodologic studies are also conducted to improve exposure assessment techniques and to understand exposure patterns, such as peak exposures. A study is being conducted in Shanghai to evaluate assessments made from detailed occupational questionnaires with air measurements. A report describing the exposure assessment procedures for evaluating EMF exposures in a brain cancer case-control study is underway. Reports describing the occurrences of pesticides, polychlorinated biphenyls, chlorinated and aromatic solvents, wood dust, lead, cadmium, and nitrosoamines are being prepared. Detailed questionnaires have been developed for several case-control studies: a kidney cancer study in the US, a bladder cancer study in New England, a NHL study in the US, and a bladder cancer study in Spain. A comparison of assessments to pesticides is being compared to pesticide metabolites in the urine. Important determinants of exposure to 2,4-D are being evaluated using urinary levels. Data mining models are being used to extract decision rules from questionnaire responses and previously assessed exposure metrics so that the decision rules can be used in subsequent studies. Comparisons of various exposure assessments methods are being made. Exposure estimates for lead and cadmium exposure are being developed for the kidney cancer study.